Method for the computing and imaging of views of an object

Abstract

With a digital volume of data representing a non-destructive investigation conducted on a body, an operation is made for the segmentation of objects contained in this digital volume, as well as an operation to display a view of this object from a given viewpoint. To perform the segmentation, the digital volume is explored according to a mode of exploration by which a parameter is assigned to each volume element, of the volume under examination, which belongs to the segmented object. Subsequently, by acting on this parameter, it is possible to modify the segmentation criterion and, hence, the criterion for determining the segmented object, only at certain places in this object. Thus, the legibility of the displayed images is improved.

What is claimed is: 
     
       1. A method for creating and displaying views of an object, said object being determined, within a digital volume, by a set of volume elements, each volume element being loaded with at least one physical datum, a value of which represents a nature of said object, said method comprising the steps of: segmenting the digital volume by digital selecting, of volume elements which are found aligned with rays casted from an arbitrary point in space, said selected volume elements being the first volume elements met by said casted ray when going over said casted rays from a first end of each of said casted rays to a second end of each of said casted rays, and having a physical datum as part of said volume elements complying with a given criterion;   wherein said step of segmenting further comprises the step of assigning each of said segmented volume elements a parameter representing one of said casted rays reaching this volume element; and   displaying an image of a view, from a given viewpoint, of said segmented object.   
     
     
       2. A method according to claim 1, whereby the selecting is conducted according to an observation mode, said casted rays starting from a given viewpoint, said given viewpoint being at the apex of an angle of observation of said observation mode, wherein an assignment of a parameter to each volume element represents coordinates of one of said casted rays, said one of said casted rays starting from said given viewpoint and reaching each of said volume elements. 
     
     
       3. A method according to claim 2 wherein the value of the observation angle is zero, the observation point being at the at infinity, to give an axonometrical representation of the segmented object. 
     
     
       4. A method according to claim 2 wherein the parameter of each element represents the coordinates of the intersection, with a reference plane, of said casted ray that reaches said element. 
     
     
       5. A method according to claims 1, 2, 3 or 4, wherein a segmentation criterion comprises: a determination of a range of values of physical data and the selection of volume elements which, in one step of the mode of exploration of the digital volume, are loaded with data, the value of which comes within this range.   
     
     
       6. A method according to claim 5 wherein a segmentation criterion further comprises: a measurement of a physical data gradient of the elements for which the values are located within the range, and a choice of the element for which this gradient is at its highest.   
     
     
       7. A method according to claim 6 wherein the representation comprises: a determination of perpendiculars to a surface of the segmented volume, and wherein said perpendiculars are directions of said gradients at the position of the corresponding volume elements.   
     
     
       8. A method according to claim 6 wherein a computation of the gradient at each element comprises: a taking into account of the elements that are closest to this element.   
     
     
       9. A method according to claim 6 wherein a computation of the gradient at each element comprises: a taking into account of a given number of elements neighbouring the element considered and a choice of an algorithm to compute this gradient.